I agree with Jim overall, but I would somewhat delineate the topic into a few sub-divisions.
Inputs and outputs:
Knowing oil lube properties (PDS) and such can speak to the propensity of any lube to succeed or fail, but UOAs speak to how well the lube actually did it's job.
Direct and indirect contributors:
Knowing how well a filter performs via a PC count certainly can tell us how well the filter cleaned the oil, but it does not tell us how well the oil protected the engine. Filtration efficiency only speaks to the filters ability to do it's direct work; that of cleaning particulate from the liquid medium. But again, UOAs tell us how well the oil did in reducing wear.
I certainly agree that a UOA is not a direct tell-tale indicator of how well a filter does in efficiency performance. But then again, filter efficiency does not tell me how well a lube does in protecting the equipment. You can have a GREAT filter, but if paired with a lousy lube, the engine (tranny, gearbox, whatever) isn't going to live long. Conversely, you could have a great lube, with poor filtration, and the equipment might be compromised as well.
Generally:
PC's speak directly to filter efficiency, and cannot speak to lube performance.
UOA's speak directly to lube performance, contamination, and other fluid characteristics.
I will acknowledge that "contamination" in a UOA is seen by many different views. We would call fuel dilution contamination. But we also call coolant ingress contamination as well. Same goes for silicon, which (if it's via air ingestion) could be particulate. Contaminiation can also be soot/insolubles.
We care about the whole system, as it operates as a package. How does the oil and filter do as a team?
So - which report (a UOA or a PC) better describes the overall team performance? In my opinion, a UOA. The UOA DIRECTLY describes and tells of wear metals. The particles in a UOA a typically small; always less than 5um in size from the perspective of spectral analysis. The vast majority of wear is predicted by small shifts that become large if neglected. If given a dead-man's choice of one or the other, I would want to know how my wear is doing in a UOA before I'd call for the PC.
Essentially, the output of a PC report (filter efficiency) is an input to the UOA process. PCs are fantastic tools for deciding which filter is better or worse than another filter; PCs speak directly to the effiency of the filter media. PCs help us decide which filter to use, relative to our goals. But PCs cannot tell us about wear; they can only predict how wear might be affected.
However, the UOA is the output telling us how well the entire system performed. UOAs do not speak to the filter directly, but they do infer some acceptable level of performance, or the wear would grossly spike. The differnce? A PC has zero ability to speak to the lube performance, but the UOA at least has an indirect ability to speak to filter performance. If I want to know how well one brand of filter compares to a different brand for some application, I'll head right to a PC report. But if I want to know how well the lube system is performing as a whole, I want a UOA.
And, most UOAs show us that filter selection has little effect in a "normal" OCI. Of all the data I've collected from many differnent sources over the last several years, I cannot find a direct relationship to predict how wear can be manipulated with a filter, using a UOA as the measuring stick.
OK - blah, blah, blah ... I realize I'm noodling this down probably too far.
In a nutshell -
Use a UOA to tell you how well the lube is working.
Use a PC to help you decide which filter to buy.
You cannot use the tools to make a direct comparison of the other characteristic. But you can infer the filter is at least doing a decent job, if the UOA is admirable.
That is my logic, anyway; YMMV.
